Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

A Target Recognition System for Small Sample Underwater Images

A technology for underwater image and target recognition, applied in character and pattern recognition, instruments, biological neural network models, etc., can solve the problems of underwater images without model training, low quality of sonar images, insufficient verification, etc., to achieve automatic Deploy and accelerate learning, save manpower and time costs, and be easy to promote

Active Publication Date: 2022-03-22
SHANDONG SHIP TECH RES INST +1
View PDF10 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, due to the changeable underwater acoustic medium, the signal transmission process is susceptible to interference, resulting in low quality sonar images of targets, high noise, and indistinct features
[0003] Due to the unknown of the underwater environment, most of the targets are small samples and rare, so the acquired target image data is very little. When processing such target images, we can only rely on the experience of researchers to operate without mature processing There is no systematic model to train underwater images in a targeted manner. When analyzing underwater targets, the image processing process is complicated and the correlation between modules is not strong. There is no mature processing method for unknown and small-sample underwater images. For reference, a lot of human resources and time resources are wasted. For these small samples and unfamiliar data, the image processing method can only be selected tentatively, and the verification is not enough, which seriously affects the target recognition efficiency.

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • A Target Recognition System for Small Sample Underwater Images
  • A Target Recognition System for Small Sample Underwater Images
  • A Target Recognition System for Small Sample Underwater Images

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0034] refer to Figure 1-9 , which is the first embodiment of the present invention, this embodiment provides a target recognition system for small-sample underwater images, including: imaging module 100, picture preprocessing module 200, deep learning module 300 and evaluation module 400; wherein the imaging module 100 is used for imaging the underwater target area, including an imaging mode selection unit 101, and the imaging mode selection unit can perform different selections of optical imaging or acoustic imaging.

[0035] The image preprocessing module 200 is connected to the imaging module 100, uses different deep learning frameworks and learning modes according to the user's selection, trains the model and monitors the whole process, including the grayscale conversion unit 201, the image binarization unit 202, and the ROI area Division unit 203, contour drawing unit 204, image noise reduction and dehazing unit 205, morphological transformation unit 206, and feature ex...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a target recognition system for small sample underwater images. The results are evaluated; the method and system integrates a very comprehensive image processing operation and deep learning model, which can help users to use different ways and different combinations to achieve underwater target training and recognition, and compare the prediction effect. The combined scheme is for reference in subsequent research, effectively solving the problem that in the target recognition research of small sample underwater images, there is no scheme to follow, no method to follow, and is limited by personal professional experience, and the present invention can achieve deep The automatic deployment and accelerated learning of the learning framework can be deployed on multiple development platforms, which is easy to promote and use, and greatly saves manpower and time costs.

Description

technical field [0001] The invention relates to the technical field of underwater target recognition, in particular to a target recognition system for small-sample underwater images. Background technique [0002] In underwater target recognition, due to the limitations of the environment, optical imaging is blocked, and acoustic imaging can only be used in long-distance directions, that is, target search is carried out with the help of sonar equipment. However, due to the changeable underwater acoustic medium, the signal transmission process is susceptible to interference, resulting in low quality sonar images of targets, high noise, and indistinct features. [0003] Due to the unknown of the underwater environment, most of the targets are small samples and rare, so the acquired target image data is very little. When processing such target images, we can only rely on the experience of researchers to operate without mature processing There is no systematic model to train und...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Patents(China)
IPC IPC(8): G06V20/00G06V10/25G06V10/30G06V10/50G06V10/764G06V10/82G06K9/62G06N3/04G06N3/08
CPCG06N3/08G06V20/00G06V10/30G06V10/25G06V10/507G06N3/045G06F18/241
Inventor 于昌利周晓滕
Owner SHANDONG SHIP TECH RES INST
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products